Abstract: In order to improve the prediction accuracy and realtime performance of the navigation satellite clock error, a satellite clock error prediction method based on the combination of polynomial model and functional network is proposed in this paper. Firstly a polynomial model is used to fit the clock error series according to its physical property, and then the polynomial fitting residual is modeled based on the functional network, while the functional network structure is defined by using wavelet denoising and phase space reconstruction. Finally the GPS satellites are taken for example and four shortterm prediction tests are done, and the simulation results show that the proposed method can be used to fit and predict the clock error series effectively, whose prediction accuracy is better than that of
IGUP ephemeris.

Key words: Clock error prediction, Functional network, Phase space reconstruction, Wavelet denoising